Where the money originates — and where the terms are dictated. OpenAI holds ~24 disclosed publisher deals; most of the rest hold one, or none.
The would-be exchanges of the licensed-content economy: they price, license, deliver, and pay. Many venues, no common rail between them.
The only place a stated preference becomes an actual gate. CDNs and bot managers decide, request by request, who gets in and at what price.
Many pens, no single page: declaration, licensing, provenance, and attribution standards are all in flight, none yet universal.
The source of everything the others trade in — and the only constituency still negotiating one contract at a time.
Prebid's home constituency — SSPs, exchanges, and publisher monetization firms. The largest bloc at our table, and the one with no seat where the AI content market is being designed.
The vendors who instrument today's open web — identity, data, engineering. Absent from the conversations shaping the AI content market; PostIndustria, which built the RAMP PoC, is the exception.
We explored the full solution space before narrowing. What survived is what the ecosystem is actually ready for.
PARSE: client-side adapter collapsing N×M marketplace integrations into N+M; marketplace-neutral; TollBit and OpenAttribution interest confirmed. RAMP: open transaction protocol; working PoC, 2 publishers adopting. Together: PARSE welcomes today's players, RAMP builds tomorrow's.
Publisher-side evidence, published: article series documenting the state of the problem with real traffic and revenue data. The instrument for getting the right people into the room.
PARSE solves the many-to-many problem for existing players today. RAMP provides the open-source path for new marketplaces tomorrow.
From "what is happening to my traffic" to "how do I evaluate any offer" — each piece useful on its own, in under 15 minutes.
The problem, from real publisher logs: automated vs human traffic, infrastructure costs, clawback risk. By Raluca Mitchell.
A reusable map for any offer: who is paid, for what use, on whose terms, through whom, with what proof.
The four monetisation models — pay-per-crawl, pay-per-use, direct deals, attribution pools — and how the money actually moves.
Marketplaces and intermediaries compared: fees, pricing, control, and reach.
robots.txt, llms.txt, RSL, CoMP — what each standard covers, what it doesn't, and which work today.
Proving where content came from and where it's used, once the click disappears.
From clicks to citations: answer share, citation frequency, GEO alongside SEO.
Copyright cases, licensing terms, and regulation shaping every deal.
Use cases for AI systems built on permissioned content and data.
* AI wasn't used to form this objectively AI-y statement